INTRODUCTION: Sarcoidosis is a rare inflammatory disease with unclear natural history. Using a large, retrospective, longitudinal, population-based cohort, we sought to define its natural history in order to guide future clinical studies. METHODS: We identified 722 newly diagnosed cases of sarcoidosis within Kaiser Permanente Northwest health care records between 1995 and 2015. We investigated immunosuppressive medication use in the two years following diagnosis, analyzed demographic and clinical characteristics, and quantified chest imaging and pulmonary function testing (PFTs) across the clinical course. RESULTS: Over two years of follow-up, 41% of patients were treated with prednisone. Of those, 75% tapered off their first course within 100 days, although half of those patients required recurrent therapy. Five percent of the entire cohort remained on prednisone for longer than one year, with an average daily dose of 10-20 mg. Chest imaging was associated with early prednisone use, and chest CT was associated with changes in prednisone dose. PFTs or demographics were not associated with prednisone use. Cumulative prednisone doses were significantly higher in African Americans (1,845 mg additional) and those who had a chest CT (2,015 mg additional). Overall, PFTs were less frequently obtained than chest imaging and had no significant change over disease course. DISCUSSION: The natural history of sarcoidosis varies greatly. For those requiring therapy, corticosteroid burden is high. Chest imaging drives medication dose changes as compared to PFTs, but neither outcome fully captures the entire history of disease. Prospective cohorts are needed with purposefully collected, repeated measures that include objective clinical assessments and symptoms.